SeedBench is the first multi-task benchmark designed to evaluate large language models (LLMs) in seed science, focusing on seed breeding. This repository includes the dataset, evaluation code, and documentation to support research in this domain.
GitHub page
Gene Information Retrieval
Gene Function and Regulation… See the full description on the dataset page:
https://huggingface.co/datasets/yj12869741/SeedBench.